Examination And Optimization On The Algorithms Of Mining Association Rules

In order to adapt the new demands of information processing, data mining has entered a new practical phase in the current data warehouse as a new information analysis technology. Data mining, also known as knowledge discovery in databases, is a process that knowledge and information of people do not know in advance and covered in large data but potentially useful, is picked up from large, noisy, incomplete, ambiguous, random data,the main purpose of which is to discover valuable knowledge from vast amounts of data for users.Mining association rules is a very important research direction of data mining. Generally speaking, the association rule mining can be divided into two steps:(1) find all frequent itemsets without omission:counts all of the frequent itemsets appeared in database must greater than or equal to a predefined minimum support;(2) the strong association rules generated from frequent itemsets:all of these rules must meet the predefined minimum confidence and the predefined minimum support. Performance of association rule mining algorithms mainly decided by the first step.In this paper, on the basis of research Apriori Algorithm which is the classical algorithm of association rules mining, analyses the advantages and disadvantages of the algorithm. To ponder over the first step of the Apriori algorithm generate a lot of candidate itemsets which are not frequent itemsets,and all of these itemsets cost a lot of system spending.This paper presents an improved algorithm DDApriori algorithm to improve the the Apriori pruning steps. Using this method, the large number of useless candidate itemsets can be reduced effectively and it can also reduce the times of judge whether the itemsets are frequent itemsets.Experimental results show that the optimized algorithm has better efficiency than classic Apriori algorithm.